Resting energy expenditure of persons with a spinal cord injury (SCI) is generally lower than that seen in able-bodied (AB) individuals due to the reduced amounts of muscle mass and sympathetic nervous system available. However, outside of clinical studies, much less data is available regarding athletes with an SCI. In order to predict the energy expenditure of persons with SCI, the generation and validation of prediction equations in relation to specific levels of SCI and training status are required. Specific prediction equations for the SCI would enable a quick and accurate estimate of energy requirements. When compared with the equivalent AB individuals, sports energy expenditure is generally reduced in SCI with values representing 30-75% of AB values. The lowest energy expenditure values are observed for sports involving athletes with tetraplegia and where the sport is a static version of that undertaken by the AB, such as fencing. As with AB sports there is a lack of SCI data for true competition situations due to methodological constraints. However, where energy expenditure during field tests are predicted from laboratory-based protocols, wheelchair ergometry is likely to be the most appropriate exercise mode. The physiological and metabolic responses of persons with SCI are similar to those for AB athletes, but at lower absolute levels. However, the underlying mechanisms pertaining to substrate utilization appear to differ between the AB and SCI. Carbohydrate feeding has been shown to improve endurance performance in athletes with generally low levels of SCI, but no data have been reported for mid to high levels of SCI or for sport-specific tests of an intermittent nature. Further research within the areas reviewed may help to bridge the gap between what is known regarding AB athletes and athletes with SCI (and other disabilities) during exercise and also the gap between clinical practice and performance.